You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
SupernovaXception model with MTS benchmark evaluation results
This PR brings in many updates, with the major additions being the inclusion of a new architecture - `snx` which is described in #58 . This PR also runs an evaluation analysis on MTS benchmark data as well as PLAsTiCC data, #52 and #44
Furthermore, with the changes made with 4dafd08 issues #54 and #56 can be closed#51 can also be closed now that there exist code in place to do comparisons.
#45 can be closed with 30f495a#39 can be closed with the inclusion of evaluation notebooks found in `astronet/notebooks`, see b7e2a84 for more.
#48 should hopefully be able to be closed depending on the outcome of this PR
It would desirable to be able to plot the weight responses of an input sequence (light curve)
This has benefits of being more explainable and would help justify the use of this mechanism for this problem.
Something akin to the image below from https://github.com/cordeirojoao/ECG_Processing/blob/master/Ecg_keras_v9-Raphael.ipynb would be useful
![image](https://user-images.githubusercontent.com/8843728/101223929-1d984200-3685-11eb-8de2-ead6ef749d9c.png)
This might be able to be achieved by branching off the
TransformerBlock
and inspecting the weights at this level. See Autoencoders notebook in https://github.com/astro-informatics/course_mlbd/blob/master/Lectures/Lecture21_AutoEncoders.ipynb for example of plotting the weightsPossible useful links:
The text was updated successfully, but these errors were encountered: